CN113933636A - Power distribution network fault test system based on arc generator - Google Patents
Power distribution network fault test system based on arc generator Download PDFInfo
- Publication number
- CN113933636A CN113933636A CN202111267561.4A CN202111267561A CN113933636A CN 113933636 A CN113933636 A CN 113933636A CN 202111267561 A CN202111267561 A CN 202111267561A CN 113933636 A CN113933636 A CN 113933636A
- Authority
- CN
- China
- Prior art keywords
- arc
- distribution network
- power distribution
- fault
- voltage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Abstract
The invention provides a power distribution network fault testing system based on an arc generator, which comprises a data analysis system, an upper computer control system, a data acquisition card, a data acquisition system, a control system, a signal feedback system, the arc generator, a switch switching system and a power distribution network simulation system, wherein the power distribution network simulation system comprises a multi-type load system and a multi-feeder power distribution network frame. The invention reproduces the arc fault in the power grid through the arc generator, the power distribution network simulation system simulates the power grid consisting of various types of loads and multiple loops, the arc generator is connected into different nodes of the power distribution network simulation system through the switch switching system to simulate the arc fault conditions of different lines and each trunk/branch in the power distribution network, meanwhile, the generation stage of the arc fault in the power distribution network is simulated by controlling the sequence of the switch action time in the switch switching system, and the arc generating device is controlled through closed-loop fuzzy control, so that the self-adaptive intelligent control can be carried out on the arc distance according to the arc voltage.
Description
Technical Field
The invention relates to the field of power distribution network fault testing, in particular to a power distribution network fault testing system based on an arc generator.
Background
Arc faults are very common in urban power distribution networks and are mainly caused by poor contact and cable aging. Once an arc fault occurs, the fault branch of the electrical line needs to be cut off at the first time to avoid further causing an electrical fire, especially the fire is a very fatal danger to the area where people gather, and the electricity quality and the personal safety are seriously influenced. The concealment and randomness of arc faults in power distribution networks is one of the difficult problems to detect such faults, which has attracted a great deal of attention from scholars. With the rapid development of power distribution networks, in recent years, a plurality of high-power test laboratories are built in a plurality of research structures and are mainly used for acquiring arc fault characteristics and verifying corresponding action characteristics and detection capability of arc fault protection electric appliances. While arc fault record data in real power systems or high power test laboratories has high confidence, these experiments require significant manpower, material resources, and financial resources to perform the tests. Since the arc fault test system in the above experiment is not changeable, the fault characteristics of arc faults of different urban distribution networks cannot be obtained. Arc fault simulation tests are an effective way to study arc fault characteristics.
With the rapid development of electric power construction, the power distribution network is more complicated than the conventional power distribution network. Due to the lack of experimental data, it is difficult to obtain accurate arc fault signatures associated with different fault conditions in a power distribution network. While existing arc generators typically integrate a simplified system thevenin equivalent circuit to study dynamic arc behavior, they cannot be used for arc fault detection and location studies of power distribution systems because they ignore the interaction between arc faults and complex power distribution systems. Therefore, for the algorithm of arc fault detection and positioning of the power distribution network, a test system formed by fusing an intelligent arc generator and a scaled-down power distribution network simulation system is required to be used for verification in a controllable experimental environment, and the arc fault generator should reproduce dynamic arc characteristics and can be conveniently connected to the scaled-down power distribution network test system to execute the transient processes of different arc faults; at the same time, the arc fault generators and the scaled down power distribution network simulation system should be controllable to apply the arc faults under remote setup commands.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a power distribution network fault testing system based on an arc generator.
A power distribution network fault testing system based on an arc generator comprises a data analysis system, an upper computer control system, a data acquisition card, a data acquisition system, a control system, a signal feedback system, the arc generator, a switch switching system and a power distribution network simulation system, wherein the power distribution network simulation system comprises a multi-type load system and a multi-feeder power distribution network frame;
the upper computer control system is used for sending a control instruction to the control system through the data acquisition card to control the switch switching system, so that the switch switching system can access the arc generator and the multi-type load system to the multi-feeder simulation power distribution network frame at different time;
the arc generator is used for reproducing arc faults occurring in the power grid;
the power grid simulation system is used for simulating a power grid consisting of multiple types of loads and multiple loops;
the data acquisition system is used for acquiring the electrode position of the arc generator in the running state, the arc voltage and the arc current between the two electrodes and the voltage and the current of each trunk and branch in the power distribution network simulation system in real time;
the signal feedback system is used for feeding back on-off state information of each switch in the switch switching system, information of a pressure sensor in the arc generator and information of front and rear limit switches of the movable electrode to the upper computer control system in real time through a data acquisition card;
the upper computer control system is also used for acquiring information in the signal feedback system and the data acquisition system through the data acquisition card, acquiring data in real time and monitoring the operation state of the system, calculating a proper electrode moving speed through a fuzzy reasoning principle according to the difference value between the acquired arc voltage feedback value and a preset voltage and the arc gap distance, and further controlling the electrode moving speed in the arc generator to enable the arc gap distance to be subjected to self-adaptive dynamic adjustment along with the arc burning state;
the data analysis system is used for analyzing the voltage and current data of each line in the power distribution network simulation system acquired from the upper computer control system so as to output fault information, and analyzing the arc faults existing in the power distribution network fault test system based on the arc generator by using the fault analysis module and outputting the fault information.
Furthermore, the control system comprises a switch control module and a stepping motor control module, the upper computer control system sends a control instruction to the switch control module of the control system through a data acquisition card so as to control the on-off of a switch in the switch switching system and the load switching of a multi-type load system in the power distribution network simulation system, wherein the switch in the switch switching system is used for switching the electric arc generated by the electric arc generator to different feeder lines of a multi-feeder line power distribution network frame in the power distribution network simulation system; the upper computer control system also sends a control instruction to a stepping motor control module of the control system through a data acquisition card, and controls the electrode movement of the arc generator by controlling the rotation of the motor.
Further, the arc generator comprises a movable electrode and a fixed electrode, the movable electrode and the fixed electrode are located on a base guide rail on the same straight line and are connected in series or merged into a circuit of a test network through a lead, the movable electrode controls the displacement direction and the speed of the electrode through a motor base linkage shaft, the movable electrode limits the maximum value and the minimum value of movement through a front limit switch and a rear limit switch, a pressure sensor is installed at the rear end of the fixed electrode 1 and is used for detecting whether the two electrodes are tightly closed, and a stepping motor of the arc generator is controlled by using a fuzzy control algorithm so as to adjust the arcing state.
Further, the fuzzy control algorithm is specifically operative to: collecting arc voltage by using a voltage sensor and calculating the effective value U of the voltagefWill U isfObtaining a voltage difference e as a difference between the voltage feedback quantity and the set voltageWhen the gap voltage is smaller than the lower limit of the given voltage, the value of the voltage difference e is the voltage lower limit minus the arc gap voltage; when the arc gap voltage is larger than the upper limit of the given voltage, the voltage difference e is obtained by subtracting the arc gap voltage from the upper limit of the voltage, then the voltage difference e and the arcing gap s are used as the input of a fuzzy controller, fuzzification, fuzzy reasoning and defuzzification processing are carried out on the input, the given motor speed is output, the moving direction and the speed of the electrode are changed according to the given motor speed, the size of the arcing gap is controlled, and the stable combustion of the arc is maintained in a self-adaptive mode.
Furthermore, the multi-feeder distribution network frame in the distribution network simulation system is equivalent to the basic line attribute in the distribution network simulation system in a series-parallel connection combination mode of a resistance load, an inductance load and a capacitance load, the multi-type load system comprises load devices of a motor type, a complete set of electrical and automation devices, the upper computer control system sends a control instruction to the control system through the data acquisition card, and the multi-type complex loads in the multi-type load cabinet are connected in series or merged into the multi-feeder simulation distribution network frame through the switch switching system to reproduce the operation, the access and the disconnection under the condition of different types of loads in the distribution network simulation system.
Furthermore, the data analysis system comprises an online detection module and an offline training module, wherein the online detection module is used for detecting and analyzing a test process and a test result of the power distribution network fault test system based on the arc generator, and the offline training module is used for training a convolutional neural network fault detection algorithm.
Further, the online detection module is specifically configured to: voltage and current signals in a line in a test process are acquired through a data acquisition system, time domain characteristics of the voltage and current signals, frequency domain characteristics extracted through fast Fourier transform and signal detail characteristics extracted through wavelet transform are utilized, the acquired characteristic values are detected through a multi-criterion fault detection module, results are transmitted to a multi-data fusion fault judgment module, the voltage and current signals are subjected to data processing, high latitude characteristics of arc voltage and current waveform images are respectively extracted through constructing a multi-layer convolutional neural network, the results are transmitted to the multi-data fusion fault judgment module, the multi-data fusion fault judgment module monitors arc faults in a power distribution network fault test system based on an arc generator based on convolutional neural network fault detection results and simultaneously utilizes the combination of the multi-criterion fault detection results, control instructions in a control system and feedback information in a signal feedback system as auxiliary criteria to jointly monitor the arc faults in the power distribution network fault test system based on the arc generator, and analyzing the arc fault and transmitting fault information to an upper computer control system.
Further, the offline training module is specifically configured to: the method comprises the steps of building a fault database by utilizing test data of a power distribution network fault test system based on an arc generator, carrying out normalization processing on the fault data, dividing a data set into a training set and a test set, carrying out one-dimensional convolutional neural network training on the training set, recording model accuracy in the training process, obtaining an optimal model when judging that the training is finished, continuing carrying out one-dimensional convolutional neural network training when judging that the training is not finished, simultaneously testing the obtained optimal model by utilizing the data of the test set, and finally evaluating the model
The invention has the following characteristics:
1. the invention provides an electric fire fault cause testing system combining an arc generator and a power distribution network, wherein the arc generator reproduces arc faults occurring in the power distribution network, the power distribution network simulation system simulates the power distribution network consisting of various types of loads and multiple loops, the arc generator is connected into different nodes of the power distribution network simulation system through a switch switching system to simulate the arc faults of different lines and branches in the power distribution network, and meanwhile, the occurrence stage of the arc faults in the power distribution network is simulated by controlling the sequence of switch action time in the switch switching system;
2. the fuzzy control algorithm is applied to the arc generator, in the dynamic process of arc burning of the arc generator, the fuzzy controller is designed according to the arc voltage and the arc gap distance, the difference value of the arc voltage feedback value and the preset voltage and the arc gap distance are used as the input of the fuzzy controller, and the fuzzy controller calculates the appropriate electrode moving speed according to the fuzzy reasoning principle, so that the electrode moving speed in the arc generator is controlled by a motor to enable the arc gap distance to be subjected to self-adaptive dynamic adjustment along with the arc burning state, the arc generating device is subjected to closed-loop fuzzy control, the arc burning distance can be subjected to self-adaptive intelligent control according to the arc voltage, stable burning of the arc is maintained, and the good effect of prolonging the arc burning time of a small-current resistive load is achieved;
3. the invention provides an arc fault multi-data fusion detection and analysis method applied to an electrical fire fault cause test system, which is characterized in that a multilayer convolutional neural network is constructed to respectively extract high latitude characteristics of arc voltage and current waveform images, and arc faults in the electrical fire fault cause test system are jointly detected by utilizing time domain characteristics of voltage and current signals, frequency domain characteristics extracted by using fast Fourier transform, signal detail characteristics extracted by using wavelet transform, system control instructions and feedback information as auxiliary criteria, half wave numbers of arcs in specified time are counted, whether the waveform meets the requirements of national standards is judged, and therefore whether the test is successful and whether a tested sample machine is qualified are determined.
Drawings
FIG. 1 is a schematic diagram of a power distribution network fault testing system based on an arc generator according to an embodiment of the present invention;
FIG. 2 is a schematic view of the construction of the arc generator of the present invention;
FIG. 3 is a schematic diagram of the control strategy of the arc generator of the present invention;
FIG. 4 is a schematic diagram of a power distribution network simulation system according to the present invention;
FIG. 5 is a schematic diagram of the data analysis system of the present invention;
FIG. 6 is a graph of the relationship between the number of iterations and the accuracy in detecting an arc fault in a multi-layer convolutional neural network according to an embodiment of the present invention;
FIG. 7 is a graph of the number of iterations versus the loss value in detecting an arc fault in a multi-layer convolutional neural network in accordance with an embodiment of the present invention.
In the figure: 1-a fixed electrode; 2-moving the electrode; 3-an electrode holder; 4-electrode backrest; 5-a pressure sensor; 6-electrode lead; 7-rolling the base; 8-electrode base; 9-a stepper motor; 10-automatic rotating shaft; 11-manual rotating shaft; 12-electrode base linkage shaft; 13-a base guide rail; 14-limit switch;
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Fig. 1 is a schematic structural diagram of one embodiment of the power distribution network fault testing system based on the arc generator, which includes a data analysis system, an upper computer control system, a data acquisition card, a data acquisition system, a control system, a signal feedback system, the arc generator, a switch switching system, and a power distribution network simulation system, wherein the control system includes a switch control module and a stepper motor control module.
And the upper computer control system is connected with the data acquisition card and is communicated with the whole lower computer test platform through the data acquisition card. The upper computer control system sends control instructions to the control system through the data acquisition card, all the control instructions are sent to the control system through the data acquisition card, and the control system (including the switch control module and the stepping motor control module) directly controls the switch switching system, the arc generator and the power distribution network simulation system.
Specifically, the upper computer control system sends a control instruction to a switch control module of the control system through a data acquisition card so as to control the on-off of a switch in a switch switching system and the load switching of a multi-type load system in the power distribution network simulation system, wherein the switch in the switch switching system is used for switching the electric arc generated by an electric arc generator to different feeder lines of a multi-feeder line distribution network frame in the power distribution network simulation system.
The upper computer control system also sends a control instruction to a stepping motor control module of the control system through a data acquisition card, and controls the electrode movement of the arc generator by controlling the rotation of the motor. The electric arc generator is used for reproducing electric arc faults occurring in the power grid, the power distribution network simulation system is used for simulating the power grid consisting of various types of loads and multiple loops, the electric arc generator is connected into the power distribution network simulation system through the switch switching system, the electric arc generator is connected into different nodes of the power distribution network simulation system through the switch switching system to simulate the electric arc faults of different lines and all branches/branches in the power distribution network, and meanwhile, the generation stage of the electric arc faults in the power distribution network is simulated by controlling the sequence of switch action time in the switch switching system.
Referring to fig. 2, the arc generator includes a moving electrode 2 and a fixed electrode 1, the moving electrode 2 and the fixed electrode 1 are located on a base guide rail 7 of the same straight line and are connected in series or merged into a circuit of a test network through a lead, the moving electrode 2 controls the displacement direction and speed of the electrodes through a motor base linkage shaft 12, the moving electrode 2 limits the maximum value and the minimum value of movement through a front limit switch and a rear limit switch 14, and a pressure sensor 5 is installed at the rear end of the fixed electrode 1 for detecting whether the two electrodes are closed tightly. The arc generator stepper motor 9 is controlled using a fuzzy control algorithm to adjust the arcing state. In the dynamic process of arc burning of the arc generator, a fuzzy controller (the fuzzy controller refers to the whole fuzzy control algorithm which controls a stepping motor of the arc fault generator) is designed according to the arc voltage and the arc gap distance, the difference value of the arc voltage feedback value and the preset voltage and the arc gap distance are used as the input of the fuzzy controller, the fuzzy controller calculates the appropriate electrode moving speed according to the fuzzy reasoning principle, and therefore the stepping motor 8 is used for controlling the electrode moving speed of the arc generator to enable the arc gap distance to be subjected to self-adaptive dynamic adjustment along with the arc burning state.
The data acquisition system comprises a voltage sensor, wherein the fuzzy control algorithm is specifically operative to: collecting arc voltage by using a voltage sensor and calculating the effective value U of the voltagefWill U isfThe voltage difference e is obtained as the difference between the voltage feedback quantity and the set voltage, and the set voltage is generally 25-45V. When the arc gap voltage is less than the given voltageWhen the lower limit is 25V, the voltage difference e is obtained by subtracting the arc gap voltage from the lower voltage limit; when the arc gap voltage is greater than the given voltage upper limit by 45V, the voltage difference e is obtained by subtracting the arc gap voltage from the voltage upper limit, and then the voltage difference e and the arcing gap s (the distance between the two electrodes) are used as the input of a fuzzy controller to carry out fuzzification, fuzzy reasoning, defuzzification and other processing on the fuzzy controller. The fuzzy controller outputs a given motor speed, changes the moving direction and the speed of the electrode according to the given motor speed, controls the size of the arcing clearance, and adaptively maintains the stable combustion of the arc.
Referring further to fig. 3, the fuzzy control algorithm is a multi-input single-output fuzzy controller, the input signals include the voltage difference e and the arcing gap s, when a signal is input into the fuzzy controller, the fuzzy controller fuzzifies the input signal, the fuzzification is a process of converting the physical quantity of the input signal into a corresponding language value, namely, the input signal is divided into one of 7 classes { NB, NM, NS, Z0, PS, PM and PB }, the fuzzified input signal obtains a fuzzified numerical value (the numerical value is one of NB NS Z0 PS PB) according to a fuzzy rule, the fuzzified numerical value needs to be defuzzified, namely, the NB is converted into the value of the physical quantity, the rotating speed of the stepping motor is controlled by the defuzzified value, the arcing distance is further controlled, the arc voltage is maintained within a preset range, and the arcing time is prolonged.
The fuzzy control is mainly used for solving the problems of a complex nonlinear system, difficulty in establishing an accurate mathematical model and control containing uncertain factors. The electric arc has chaotic characteristics, and the fuzzy algorithm is used for better controlling the electric arc generator to generate the electric arc which meets the test. Applying fuzzy control on the arc generating device: although the arc voltage and arc gap distance are directly proportional, there is no clear model that can describe the quantitative relationship (non-linearity) between the two. The fuzzy control is mainly used for solving the control problems that a complex nonlinear system is difficult to establish an accurate mathematical model and uncertain factors. Therefore, a fuzzy control strategy is applied to the arc voltage model by combining the characteristics of arc voltage nonlinearity and model uncertainty.
Fuzzy logic reasoning requires conversion between the amount of sharpness and the amount of blur for the input variable and the output variable, the input variable voltage difference e is divided into 5 fuzzy subsets NB, NS, Z0, PS, PB, the input variable arcing gap s and the output variable y are divided into 7 fuzzy subsets NB, NM, NS, Z0, PS, PM, PB.
Establishing a fuzzy control rule needs to consider the arcing characteristics of an arc generating device, such as: when the two electrodes start to separate, the voltage difference e is larger, and the arc gap s is smaller, which indicates that the deviation of the arc gap voltage and the target voltage range is larger, and the arcing interval is increased at a larger speed.
Performing fuzzy reasoning by adopting a Mamdani algorithm of parallel computing according to the established fuzzy rule, wherein an implication relation matrix R is as follows:
fuzzy rule table
The area center method with smooth output, simple calculation and high precision is adopted for defuzzification: firstly, calculating the area of the converted membership function in the output variable range, and then calculating the geometric center of the area by using the following formula, wherein the geometric center is used as the optimal value of fuzzy inference:
wherein u iscenIs the center covered by the membership μ u (z) function of the fuzzy set u.
Referring to fig. 4, the power distribution network simulation system is constructed according to the actual typical power distribution network in a scaling-down manner, and comprises a multi-type load system and a multi-feeder power distribution network frame. The multi-feeder distribution network frame is equivalent to the basic line attribute in the distribution network simulation system in a series-parallel connection combination mode of a resistance load, an inductance load and a capacitance load. The multi-type load system comprises a plurality of types of load devices such as motors, complete sets of electric and automatic devices and the like. The upper computer control system sends a control instruction to the control system through the data acquisition card, and the multi-type complex loads in the multi-type load cabinet are connected in series or merged into the multi-feeder simulation power distribution network frame through the switch switching system to reproduce the operation, connection and disconnection under the condition of different types of loads in the power distribution network simulation system.
The data acquisition system is used for acquiring the electrode position of the arc generator in the running state, the arc voltage and the arc current between the two electrodes and the voltage and the current of each trunk and branch in the power distribution network simulation system in real time. And the signal feedback system is used for feeding back on-off state information of each switch in the switch switching system, pressure sensors in the arc generator and information of front and rear limit switches of the movable electrode to the upper computer control system in real time through a data acquisition card. The upper computer control system acquires information in the signal feedback system and the data acquisition system through the data acquisition card to acquire data in real time and monitor the running state of the system. The signal feedback system feeds back switch on-off information (through a contactor auxiliary contact) and limit switch information in the switch switching system, comprises optical couplers, a shift register and other components, is a mature product, and has a large amount of applications in PLC and other applications.
The data analysis system is used for analyzing voltage and current data of each line in the power distribution network simulation system acquired from the upper computer control system so as to output fault information, analyzing arc faults existing in the power distribution network fault test system based on the arc generator by using the fault analysis module, and outputting the fault information, wherein the arc fault information comprises the steps of counting arc half wave numbers in specified time, judging whether a waveform meets corresponding national standards or test requirements, and determining whether a test is successful or not and whether a tested sample machine is qualified or not. The method comprises the steps of collecting and classifying various fault waveforms, establishing a typical waveform database, and establishing a detection and analysis system integrating waveform collection, data processing, feature analysis and performance judgment into a whole through waveform processing and analysis.
As shown in fig. 5, the data analysis system includes an online detection module and an offline training module.
The offline training module is used for training the convolutional neural network fault detection algorithm. The method comprises the steps of building a fault database by using test data of a power distribution network fault test system based on an arc generator, carrying out normalization processing on the fault data, dividing a data set into a training set and a test set, wherein the training set is used for carrying out one-dimensional convolutional neural network training, recording model accuracy in the training process, obtaining an optimal model when judging that the training is finished, continuing the one-dimensional convolutional neural network training when judging that the training is not finished, simultaneously testing the obtained optimal model by using the data of the test set, and finally evaluating the model.
The online detection module is used for detecting and analyzing the test process and the test result of the power distribution network fault test system based on the arc generator. Voltage and current signals in a line in a test process are acquired through a data acquisition system, the time domain characteristics of the voltage and current signals, the frequency domain characteristics extracted through fast Fourier transform and the signal detail characteristics extracted through wavelet transform are utilized, the acquired characteristic values are detected through a multi-criterion fault detection module, and the result is transmitted to a multi-data fusion fault judgment module. And carrying out data processing on the voltage and current signals, respectively extracting high latitude characteristics of the arc voltage and current waveform images by constructing a multilayer convolutional neural network, and transmitting the result to a multi-data fusion fault judgment module. The multi-data fusion fault judgment module monitors the arc fault in the power distribution network fault test system based on the arc generator by using the multi-criterion fault detection result, the control instruction in the control system and the feedback information in the signal feedback system in combination as auxiliary criteria based on the fault detection result of the convolutional neural network, analyzes the arc fault and transmits the fault information to the upper computer control system.
The multi-data fusion fault judgment module is used for making up for the defects in single detection by reasonably utilizing the time domain, frequency domain and time-frequency domain characteristics of the fault; meanwhile, in order to eliminate the subjectivity problem existing in the traditional current signal fault feature extraction process, the waveform of the fault signal is directly input into a fault arc detection model so as to release the feature expression capability of the arc fault current, and the convolution neural network is utilized to detect and classify the fault so as to improve the maximization of fault detection and data analysis.
The invention has the following characteristics:
1. an electrical fire fault cause test system is presented that combines an arc generator and a power distribution network. The arc generator is used for reproducing arc faults occurring in the power grid, the power distribution network simulation system is used for simulating the power grid consisting of various types of loads and multiple loops, the arc generator is connected into different nodes of the power distribution network simulation system through the switch switching system to simulate the arc faults of different lines and all branches/branches in the power distribution network, and meanwhile, the generation stage of the arc faults in the power distribution network is simulated by controlling the sequence of switch action time in the switch switching system.
2. An arc generator electrode motion control method based on a fuzzy logic controller is provided. The present invention applies a fuzzy control algorithm to the arc generator. In the dynamic process of arc burning of the arc generator, a fuzzy controller is designed according to arc voltage and arc gap distance, the difference value of the arc voltage feedback value and preset voltage and the arc gap distance are used as the input of the fuzzy controller, the fuzzy controller calculates proper electrode moving speed according to a fuzzy reasoning principle, and therefore the electrode moving speed in the arc generator is controlled by a motor to enable the arc gap distance to be subjected to self-adaptive dynamic adjustment along with the arc burning state. The closed-loop fuzzy control arc generating device can perform self-adaptive intelligent control on the arcing interval according to the arcing voltage, maintain stable combustion of the arc, and has a good effect on prolonging the arcing time of the low-current resistive load.
3. The method is applied to detecting and analyzing arc fault multi-data fusion of an electrical fire fault cause testing system. The method comprises the steps of constructing a multilayer convolutional neural network, extracting high latitude characteristics of arc voltage and current waveform images respectively, detecting arc faults in an electrical fire fault cause testing system by using time domain characteristics of voltage and current signals, frequency domain characteristics extracted by using fast Fourier transform, signal detail characteristics extracted by using wavelet transform, system control instructions and feedback information as auxiliary criteria, counting arc half wave numbers in specified time, and judging whether waveforms meet the requirements of national standards, so that whether the test is successful and whether a tested sample machine is qualified are determined.
The invention increases the detection method result of the multilayer convolutional neural network, and meanwhile, the rapidity and the accuracy can be improved by combining a plurality of methods. In the process of detecting the arc fault by the multilayer convolutional neural network, the accuracy is gradually improved along with the increase of the iteration times. When the number of iterations reaches 90, the accuracy begins to converge and stabilizes at 99%, as shown in fig. 6. In the iteration process, the loss value shows a descending trend, and when the iteration is carried out for 95 times, the loss value is basically stabilized at 0.07, as shown in fig. 7. Time domain, frequency domain features, and signal detail features extracted using wavelet transforms have been used for arc fault detection, respectively, but have certain drawbacks. The detection result of the multilayer convolutional neural network is combined with the characteristics, the system control instruction and the feedback information, so that the rapidity and the accuracy of detection can be further ensured, namely, the fault information is output only when most of the judgment is effective. With the multilayer convolutional neural network, in the case where a fault is determined as described above, the number of fault half waves and the fault time can be detected by directly inputting the waveform of the fault signal to the fault arc detection model.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (8)
1. The utility model provides a distribution network fault test system based on arc generator which characterized in that: the power distribution network simulation system comprises a data analysis system, an upper computer control system, a data acquisition card, a data acquisition system, a control system, a signal feedback system, an arc generator, a switch switching system and a power distribution network simulation system, wherein the power distribution network simulation system comprises a multi-type load system and a multi-feeder power distribution network frame;
the upper computer control system is used for sending a control instruction to the control system through the data acquisition card to control the switch switching system, so that the switch switching system can access the arc generator and the multi-type load system to the multi-feeder simulation power distribution network frame at different time;
the arc generator is used for reproducing arc faults occurring in the power grid;
the power grid simulation system is used for simulating a power grid consisting of multiple types of loads and multiple loops;
the data acquisition system is used for acquiring the electrode position of the arc generator in the running state, the arc voltage and the arc current between the two electrodes and the voltage and the current of each trunk and branch in the power distribution network simulation system in real time;
the signal feedback system is used for feeding back on-off state information of each switch in the switch switching system, information of a pressure sensor in the arc generator and information of front and rear limit switches of the movable electrode to the upper computer control system in real time through a data acquisition card;
the upper computer control system is also used for acquiring information in the signal feedback system and the data acquisition system through the data acquisition card, acquiring data in real time and monitoring the operation state of the system, calculating a proper electrode moving speed through a fuzzy reasoning principle according to the difference value between the acquired arc voltage feedback value and a preset voltage and the arc gap distance, and further controlling the electrode moving speed in the arc generator to enable the arc gap distance to be subjected to self-adaptive dynamic adjustment along with the arc burning state;
the data analysis system is used for analyzing the voltage and current data of each line in the power distribution network simulation system acquired from the upper computer control system so as to output fault information, and analyzing the arc faults existing in the power distribution network fault test system based on the arc generator by using the fault analysis module and outputting the fault information.
2. The arc generator based power distribution network fault testing system of claim 1, wherein: the control system comprises a switch control module and a stepping motor control module, the upper computer control system sends a control instruction to the switch control module of the control system through a data acquisition card so as to control the on-off of a switch in the switch switching system and the load switching of a multi-type load system in the power distribution network simulation system, wherein the switch in the switch switching system is used for switching the electric arc generated by the electric arc generator to different feeder lines of a multi-feeder line power distribution network frame in the power distribution network simulation system; the upper computer control system also sends a control instruction to a stepping motor control module of the control system through a data acquisition card, and controls the electrode movement of the arc generator by controlling the rotation of the motor.
3. The arc generator based power distribution network fault testing system of claim 1, wherein: the electric arc generator comprises a movable electrode and a fixed electrode, wherein the movable electrode and the fixed electrode are positioned on a base guide rail in the same straight line and are connected in series or merged into a circuit of a test network through a lead, the movable electrode controls the displacement direction and the speed of the electrode through a motor base linkage shaft, the movable electrode limits the maximum value and the minimum value of movement through a front limit switch and a rear limit switch, a pressure sensor is installed at the rear end of the fixed electrode 1 and is used for detecting whether the two electrodes are tightly closed, and a stepping motor of the electric arc generator is controlled by using a fuzzy control algorithm so as to adjust the arcing state.
4. The arc generator based power distribution network fault testing system of claim 3, wherein: the fuzzy control algorithm is specifically operative to: collecting arc voltage by using a voltage sensor and calculating the effective value U of the voltagefWill U isfWhen the arc gap voltage is smaller than the lower limit of the given voltage, the value of the voltage difference e is obtained by subtracting the arc gap voltage from the lower limit of the voltage; when the arc gap voltage is larger than the upper limit of the given voltage, the voltage difference e is obtained by subtracting the arc gap voltage from the upper limit of the voltage, then the voltage difference e and the arcing gap s are used as the input of a fuzzy controller, fuzzification, fuzzy reasoning and defuzzification processing are carried out on the input, the given motor speed is output, the size of the arcing gap is controlled according to the moving direction and the speed of the electrode which are changed according to the given motor speed, and the stable combustion of the arc is maintained in a self-adaptive mode。
5. The arc generator based power distribution network fault testing system of claim 1, wherein: the multi-feeder power distribution network frame in the power distribution network simulation system is equivalent to the basic line attribute in the power distribution network simulation system in a series-parallel combination mode of a resistance load, an inductance load and a capacitance load, a multi-type load system comprises load devices of motors, complete sets of electrical and automation devices, an upper computer control system sends a control instruction to a control system through a data acquisition card, and multi-type complex loads in the multi-type load cabinet are connected in series or merged into the multi-feeder power distribution network frame through a switch switching system to reproduce the operation, access and disconnection under different types of load conditions in the power distribution network simulation system.
6. The arc generator based power distribution network fault testing system of claim 1, wherein: the data analysis system comprises an online detection module and an offline training module, wherein the online detection module is used for detecting and analyzing a test process and a test result of the power distribution network fault test system based on the arc generator, and the offline training module is used for training a convolutional neural network fault detection algorithm. .
7. The arc generator based power distribution network fault testing system of claim 6, wherein: the online detection module is specifically configured to: voltage and current signals in a line in a test process are acquired through a data acquisition system, time domain characteristics of the voltage and current signals, frequency domain characteristics extracted through fast Fourier transform and signal detail characteristics extracted through wavelet transform are utilized, the acquired characteristic values are detected through a multi-criterion fault detection module, results are transmitted to a multi-data fusion fault judgment module, the voltage and current signals are subjected to data processing, high latitude characteristics of arc voltage and current waveform images are respectively extracted through constructing a multi-layer convolutional neural network, the results are transmitted to the multi-data fusion fault judgment module, the multi-data fusion fault judgment module monitors arc faults in a power distribution network fault test system based on an arc generator based on convolutional neural network fault detection results and simultaneously utilizes the combination of the multi-criterion fault detection results, control instructions in a control system and feedback information in a signal feedback system as auxiliary criteria to jointly monitor the arc faults in the power distribution network fault test system based on the arc generator, and analyzing the arc fault and transmitting fault information to an upper computer control system.
8. The arc generator based power distribution network fault testing system of claim 6, wherein: the offline training module is specifically configured to: the method comprises the steps of building a fault database by using test data of a power distribution network fault test system based on an arc generator, carrying out normalization processing on the fault data, dividing a data set into a training set and a test set, wherein the training set is used for carrying out one-dimensional convolutional neural network training, recording model accuracy in the training process, obtaining an optimal model when judging that the training is finished, continuing the one-dimensional convolutional neural network training when judging that the training is not finished, simultaneously testing the obtained optimal model by using the data of the test set, and finally evaluating the model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111267561.4A CN113933636B (en) | 2021-10-29 | 2021-10-29 | Power distribution network fault test system based on arc generator |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111267561.4A CN113933636B (en) | 2021-10-29 | 2021-10-29 | Power distribution network fault test system based on arc generator |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113933636A true CN113933636A (en) | 2022-01-14 |
| CN113933636B CN113933636B (en) | 2022-09-02 |
Family
ID=79284837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111267561.4A Active CN113933636B (en) | 2021-10-29 | 2021-10-29 | Power distribution network fault test system based on arc generator |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113933636B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080100305A1 (en) * | 2006-11-01 | 2008-05-01 | Eaton Corporation | Automated arc generator and method to repeatably generate electrical arcs for AFCI testing |
| CN103698677A (en) * | 2014-01-17 | 2014-04-02 | 福州大学 | Low-voltage arc fault test and analysis system |
| CN105425118A (en) * | 2015-10-29 | 2016-03-23 | 山东建筑大学 | Multi-information fusion fault arc detection method and device |
| CN105743075A (en) * | 2016-04-22 | 2016-07-06 | 国家电网公司 | Single-phase arc-grounding simulation device for power distribution network |
| CN106825953A (en) * | 2017-01-22 | 2017-06-13 | 大连理工大学 | A kind of hybrid Laser-Arc Welding real-time monitoring system and its regulation and control method |
| CN107765150A (en) * | 2017-10-18 | 2018-03-06 | 福州大学 | Intelligent electric arc fault simulation system and operating method |
| CN108963962A (en) * | 2018-08-10 | 2018-12-07 | 中航建设集团成套装备股份有限公司 | A kind of multi-layer arc fault open-circuit system |
| CN109142851A (en) * | 2018-07-26 | 2019-01-04 | 福州大学 | A kind of novel power distribution network internal overvoltage recognition methods |
| CN110488161A (en) * | 2019-07-23 | 2019-11-22 | 南京航空航天大学 | A kind of detection of multi-load series arc faults and localization method |
| CN111458599A (en) * | 2020-04-16 | 2020-07-28 | 福州大学 | Series arc fault detection method based on one-dimensional convolutional neural network |
| CN111707904A (en) * | 2020-06-17 | 2020-09-25 | 华中科技大学 | Distribution network physical simulation experiment system with arc light grounding variable structure |
| CN112701883A (en) * | 2020-12-29 | 2021-04-23 | 上海电机学院 | Power grid simulator control system and method based on fuzzy PI and QPR |
-
2021
- 2021-10-29 CN CN202111267561.4A patent/CN113933636B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080100305A1 (en) * | 2006-11-01 | 2008-05-01 | Eaton Corporation | Automated arc generator and method to repeatably generate electrical arcs for AFCI testing |
| CN103698677A (en) * | 2014-01-17 | 2014-04-02 | 福州大学 | Low-voltage arc fault test and analysis system |
| CN105425118A (en) * | 2015-10-29 | 2016-03-23 | 山东建筑大学 | Multi-information fusion fault arc detection method and device |
| CN105743075A (en) * | 2016-04-22 | 2016-07-06 | 国家电网公司 | Single-phase arc-grounding simulation device for power distribution network |
| CN106825953A (en) * | 2017-01-22 | 2017-06-13 | 大连理工大学 | A kind of hybrid Laser-Arc Welding real-time monitoring system and its regulation and control method |
| CN107765150A (en) * | 2017-10-18 | 2018-03-06 | 福州大学 | Intelligent electric arc fault simulation system and operating method |
| CN109142851A (en) * | 2018-07-26 | 2019-01-04 | 福州大学 | A kind of novel power distribution network internal overvoltage recognition methods |
| CN108963962A (en) * | 2018-08-10 | 2018-12-07 | 中航建设集团成套装备股份有限公司 | A kind of multi-layer arc fault open-circuit system |
| CN110488161A (en) * | 2019-07-23 | 2019-11-22 | 南京航空航天大学 | A kind of detection of multi-load series arc faults and localization method |
| CN111458599A (en) * | 2020-04-16 | 2020-07-28 | 福州大学 | Series arc fault detection method based on one-dimensional convolutional neural network |
| CN111707904A (en) * | 2020-06-17 | 2020-09-25 | 华中科技大学 | Distribution network physical simulation experiment system with arc light grounding variable structure |
| CN112701883A (en) * | 2020-12-29 | 2021-04-23 | 上海电机学院 | Power grid simulator control system and method based on fuzzy PI and QPR |
Non-Patent Citations (2)
| Title |
|---|
| WANHAI WU ET AL.: "《Design of Arc Generator Based on Electromechanical Coupling Closed-Loop Fuzzy Control》", 《PROCEEDINGS OF 2021 CHINESE INTELLIGENT SYSTEMS CONFERENCE. LECTURE NOTES IN ELECTRICAL ENGINEERING》 * |
| 苏晶晶 等: "《基于EMD和PNN故障电弧多变量判据诊断方法》", 《电力自动化设备》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113933636B (en) | 2022-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Amraee et al. | Transient instability prediction using decision tree technique | |
| CN102074955B (en) | Method based on knowledge discovery technology for stability assessment and control of electric system | |
| CN109948597B (en) | High-voltage circuit breaker mechanical fault diagnosis method | |
| CN110829417B (en) | Electric power system transient stability prediction method based on LSTM double-structure model | |
| Azmy et al. | Artificial neural network-based dynamic equivalents for distribution systems containing active sources | |
| Yang et al. | Condition evaluation for opening damper of spring operated high-voltage circuit breaker using vibration time-frequency image | |
| CN111553495A (en) | Small circuit breaker fault analysis method based on probabilistic neural network | |
| CN112330165B (en) | Power grid transient stability evaluation method and system based on feature separation type neural network | |
| CN105938578A (en) | Large-scale photovoltaic power station equivalent modeling method based on clustering analysis | |
| CN111308260B (en) | Electric energy quality monitoring and electric appliance fault analysis system based on wavelet neural network and working method thereof | |
| CN105425768A (en) | Electric power secondary equipment monitoring device and method | |
| CN112200694B (en) | Dominant instability mode identification model construction and application method based on graph neural network | |
| Izadi et al. | A synchronized lissajous-based method to detect and classify events in synchro-waveform measurements in power distribution networks | |
| Yang et al. | Analysis of power system transient stability characteristics with the application of massive transient stability simulation data | |
| CN112564107A (en) | Transient stability assessment method for power system | |
| Ibrahim et al. | Novel data compression technique for power waveforms using adaptive fuzzy logic | |
| Ndjakomo Essiane et al. | Faults detection and identification in PV array using kernel principal components analysis | |
| CN113094983B (en) | Online simulation method for multi-dimensional time-varying characteristics of direct-current fault electric arc of photovoltaic system | |
| CN111062569A (en) | Low-current fault discrimination method based on BP neural network | |
| Fornás et al. | Detection and classification of fault types in distribution lines by applying contrastive learning to GAN encoded time-series of pulse reflectometry signals | |
| CN113933636B (en) | Power distribution network fault test system based on arc generator | |
| Niazazari et al. | Event cause analysis in distribution networks using synchro waveform measurements | |
| CN115712871A (en) | Power electronic system fault diagnosis method combining resampling and integrated learning | |
| CN112651628A (en) | Power system transient stability evaluation method based on capsule neural network | |
| Zhou et al. | Transient stability assessment of large-scale AC/DC hybrid power grid based on separation feature and deep belief networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |